live wire
IBM makes watsonx Orchestrate AgentOps, custom LLM judging and Bedrock-agent discovery generally availableIBMSchemaGate 0.1.45 fixes broken Oracle ADB wallet connections and an OCI stack pinned 28 releases behindSchemaGatePDI’s Amazon Quick procurement agent grounds spend answers in vendor, category and contract contextAWS Business Intelligence BlogBigQuery’s ML.METRICS example returns 0.84 accuracy but 0.30 macro-F1 on the same 100-row classification queryGoogle Cloud BigQuery docsSchemaGate 0.1.44 auto-selects sentence embeddings, lifting bundled-schema retrieval from 90/98 to 93/98SchemaGateSchemaGate 0.1.43 adds read-only SQL execution with per-principal table checks—and documents unauthenticated client assertionsSchemaGateDatabox adds reusable AI Analyst Skills with personal/company scope, auto-matching and marketplace installsDataboxFabric previews an AI builder for data-agent instructions, source guidance and example queriesMicrosoft FabricDatabricks trains data-agent retriever to stop early or spend bounded extra search steps, reporting 5.8-second latencyDatabricksThoughtSpot adds SpotterCode coding agent to its Visual Embed PlaygroundThoughtSpotLongMemEval-S audit: 67–73% of restore-fixable 80k-budget errors came from evicted evidence under three policiesarXivSchemaGate 0.1.42 adds dimension-aware retrieval and fixes complex multi-table SQL promptsSchemaGateSnowflake agent toolsets can silently drop inherited tools when callers lack accessSnowflake DocumentationLooker’s VS Code extension reaches GA with MCP-assisted LookML generation, editing and validationGoogle Cloud Looker release docsIBM makes watsonx Orchestrate AgentOps, custom LLM judging and Bedrock-agent discovery generally availableIBMSchemaGate 0.1.45 fixes broken Oracle ADB wallet connections and an OCI stack pinned 28 releases behindSchemaGatePDI’s Amazon Quick procurement agent grounds spend answers in vendor, category and contract contextAWS Business Intelligence BlogBigQuery’s ML.METRICS example returns 0.84 accuracy but 0.30 macro-F1 on the same 100-row classification queryGoogle Cloud BigQuery docsSchemaGate 0.1.44 auto-selects sentence embeddings, lifting bundled-schema retrieval from 90/98 to 93/98SchemaGateSchemaGate 0.1.43 adds read-only SQL execution with per-principal table checks—and documents unauthenticated client assertionsSchemaGateDatabox adds reusable AI Analyst Skills with personal/company scope, auto-matching and marketplace installsDataboxFabric previews an AI builder for data-agent instructions, source guidance and example queriesMicrosoft FabricDatabricks trains data-agent retriever to stop early or spend bounded extra search steps, reporting 5.8-second latencyDatabricksThoughtSpot adds SpotterCode coding agent to its Visual Embed PlaygroundThoughtSpotLongMemEval-S audit: 67–73% of restore-fixable 80k-budget errors came from evicted evidence under three policiesarXivSchemaGate 0.1.42 adds dimension-aware retrieval and fixes complex multi-table SQL promptsSchemaGateSnowflake agent toolsets can silently drop inherited tools when callers lack accessSnowflake DocumentationLooker’s VS Code extension reaches GA with MCP-assisted LookML generation, editing and validationGoogle Cloud Looker release docs
nl2sql.ai
analysisANALYSIS

Alation’s ontology layer is a governance preview—not yet a runtime guarantee

The useful architecture is a continuous loop from critical-data classification to quality, curation and agent context. Buyers still need proof that policy survives execution.

Enterprise governance loop from catalog to AI agent.
AI-generated illustration
By The News Desk· Sep 10, 2026the quick take — two AI hosts go live when you do

Alation has sketched a more complete answer to a problem enterprise data-agent teams keep running into: governing a table is not the same as governing the context an agent uses to interpret it. In a September 10 product essay, the company describes a connected loop spanning Critical Data Manager, Data Quality, Curation Automation and a forthcoming ontology layer.

What Alation says is connected

The first three pieces already form a recognizable control chain. Critical Data Manager identifies business-critical assets and maps them to obligations and processes. Data Quality suggests and monitors checks for those assets. Curation Automation uses catalog context, source comments, query history and field-level instructions to propose titles, descriptions, classifications, PII tags and steward assignments. Alation says existing steward-entered values are preserved, suggestions can be previewed, and changes are recorded in the same workflow.

That matters because an analytics agent does not consume only rows and columns. It also consumes definitions, metric relationships, policies and operational documents. If those layers drift independently, a certified table can still produce an answer based on stale business meaning.

The ontology is the important—but unproven—part

Alation says it will extend this loop with ontologies shown at revAlation Chicago on September 17. The company describes them as continuously updated, scoped maps connecting data, metrics and policies within a business process, built on open standards. It also says policy constraints will be enforced during agent interactions and traced to the source policy at runtime.

Those are stronger claims than cataloging metadata. They imply that policy context can travel with an agent request and remain inspectable after execution. But the post is a product preview, not evidence of a generally available enforcement boundary. It does not publish supported agent runtimes, an API contract, failure behavior when policy context is unavailable, or an evaluation showing that the resulting answers obey the declared rules.

What buyers should ask to see

A serious evaluation should test four handoffs:

  1. A policy change updates the ontology without manual duplication.
  2. The agent receives only the context allowed for the requesting identity and business process.
  3. A stale or missing rule fails closed rather than letting the model improvise.
  4. The answer record identifies the policy version, metadata and quality signals used.

Alation’s architecture is useful because it treats agent context as governed operational data rather than prompt decoration. The next step is proof: a runnable demonstration showing that the declared policy survives every handoff from catalog to agent response.

Filed by The News Desk. Corrections: desk@nl2sql.ai · Our standards →

comments · 0

    Comments are moderated before they appear. Your email is used once to confirm it is you — never shown, never sold. Corrections and questions get an answer from the desk when we have one.